I'm looking to make a dipole for 10 M. TX power won't be more than 10 W, likely less than 1 initially, but I would prefer that it be reasonably good performance for reception.


After a discussion in the hamshack, the idea of a planar zig-zag was suggested. This article enter link description here seems to indicate that this is a pretty well understood, and adequately efficient approach. Plus, it details how to design the zig-zag for best efficiency. So, while I might still experiment with the loose helix, this will probably be my first experiment.

For various reasons I would like to fit this in a 16 foot length. I believe the target length would be about 8 foot 4 inches for each element. My question is, if I make these elements slightly helical, so as to shorten the physical length by 4 inches per side will this badly damage the quality of the antenna? (I suppose I might quantify that by asking "will that be worse than using loading coils?"

Part two, clearly this kind of shape would need some kind of former/core (the wire itself won't hold that shape!) Is the material I use (non-metallic, likely wood) going to give me trouble?

Thanks for any input

  • 2
    $\begingroup$ Sounds like you are trying to reinvent the broadside helical antenna. Very common, works great, but the length you need might not be what you expect. Unfortunately eznec and 4nec2 are not able to model this accurately. $\endgroup$
    – user10489
    Commented Nov 17, 2021 at 0:45
  • $\begingroup$ Oh, well, I'll go investigate that, thank you. $\endgroup$ Commented Nov 17, 2021 at 1:13
  • $\begingroup$ The old nec solvers only model linear elements, it can't handle curves, and if you make lots of small segments to simulate a curve it blows up. Someone wrote a paper on exact modeling of a helix and that has been incorporated into some of the commercial solvers, but not the old nec solvers. $\endgroup$
    – user10489
    Commented Nov 17, 2021 at 23:16
  • $\begingroup$ I have EZNEC 5.0+, and it models helix antennas. But I can't install it easily (media compatibility issue). BTW, EZNEC will be free in 2022. :) $\endgroup$ Commented Nov 18, 2021 at 13:35
  • $\begingroup$ EZNEC and 4nec2 both use the free NEC-2 engine by default @MikeWaters. There are updated versions of the engine, the latest being NEC-5, but they aren't free. I question that the engine "blows up" simulating a helical antenna when EZNEC will help you model one, but I know the engine has limitations that lead to some simulation results being inaccurate. $\endgroup$
    – rclocher3
    Commented Nov 18, 2021 at 15:00

2 Answers 2


For perspective on this topic: in his textbook ANTENNAS For All Applications, 3rd Edition (McGraw-Hill), John Kraus shows in the text surrounding his Figure 8-72 that a conductor wound in the form of a normal-mode helix has very nearly the same performance characteristics as a linear conductor having the same, overall, end-end length as the helix.

However that means its radiation resistance has been reduced from that of its linear form, which may have a significant and negative affect on the gain (efficiency) of an antenna system using such coiled conductors in all or part of its construction.

  • $\begingroup$ Oh, so, I wouldn't be able to save any length on the overall structure. Good to know! $\endgroup$ Commented Nov 26, 2021 at 23:52

I would guess that slightly-helical dipole elements wouldn't give you enough inductance to compensate for the capacitance of your shortened dipole, so in effect you would have a shortened dipole. Since it would only be slightly shortened, this wouldn't be the end of the world, but you would probably need a transmatch (antenna tuner) to keep the transmitter happy. I would think that the small inductors inside a transmatch would be less efficient than loading coils, and the higher SWR on the feed line would induce higher losses there also. So unless your slightly-helical elements provide enough inductance to keep the SWR low enough to not need a transmatch, then I would say that they would be worse than loading coils. You probably wouldn't notice the loss of efficiency of a few percent, but you would certainly notice needing to use a transmatch.

Regarding using a form, I'd say that the material wouldn't matter much, as long as it's non-conductive. Wood might be OK, but it might soak up water and become heavy and somewhat conductive. I'd recommend using PVC pipe instead, if you go with the helical idea. Black PVC pipe probably resists UV better than white or gray PVC.

But rather than guessing if it would work, you might try modeling the antenna, if you're into that kind of thing. If you know how long your antenna elements will be, then you could model the antenna in EZNEC or 4nec2 as a shortened dipole made with straight wires over typical ground and get a good idea of what its impedance would be, and how much inductance you would need to have a low SWR. (I'm not sure how accurate a model of a helical antenna would be with the free NEC2 engine.) There's a well-known formula for the self-inductance of a coil of wire, and many online calculators, but I would think that the traditional formula only works for tightly-wound coils and not loose helixes. This calculator by Serge Y. Stroobandt ON4AA uses several correction factors and cites references, so perhaps it's more accurate. You could always try using it to see how tight it says your helixes would need to be to get the necessary inductance.

If you do build the antenna, please do come back here and answer your own question to tell us how it works!

  • 1
    $\begingroup$ Interesting, thank you. I had read that one can bend dipole elements (perhaps to L shapes, but presumably not with tight bends) so thought the "electrical length" would be essentially the same. Ah well, this is complex stuff, and I have lots to learn. $\endgroup$ Commented Nov 16, 2021 at 17:09
  • 2
    $\begingroup$ You're welcome! The "antenna will behave the same if the amount of wire is the same" theory, meaning that a shortened dipole will be resonant if it has loading coils the same length as the dipole is short, is clearly an oversimplification. But it might be true enough in this case; I don't know, because I haven't modeled or tried your idea. The good news is that this stuff is fun (at least for me)! $\endgroup$
    – rclocher3
    Commented Nov 16, 2021 at 18:40
  • $\begingroup$ "I would guess"? $\endgroup$ Commented Nov 18, 2021 at 14:07

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .